急性脑出血患者脑源性多器官功能障碍综合征风险预测模型的建立
该研究回顾性分析2019年1月至2023年6月皖南康复医院收治的93例急性脑出血患者,按住院期间是否发生脑源性多器官功能障碍综合征(CMODS)分为CMODS组(26例)和非CMODS组(67例)。多因素Logistic回归显示糖尿病(OR=3.615)、APACHEⅡ评分(OR=4.697)、内毒素血症(OR=4.577)和脑出血量(OR=4.039)为CMODS危险因素,据此建立列线图预测模型。模型C指数0.804,校准曲线拟合良好(P>0.05),ROC曲线AUC为0.855,敏感度76.92%、特异度86.57%。作者认为该模型对评估急性脑出血患者CMODS风险有效。
要点
- 93例急性脑出血患者中26例发生CMODS,发生率27.96%,非CMODS组67例。
- 多因素分析:糖尿病OR=3.615(95%CI 1.487-8.785,P=0.000)、APACHEⅡ评分OR=4.697(95%CI 1.933-11.416,P=0.000)、内毒素血症OR=4.577(95%CI 1.883-11.123,P=0.000)、脑出血量OR=4.039(95%CI 1.662-9.816,P=0.000)。
- CMODS组糖尿病比例、APACHEⅡ≥35分比例、脑出血量≥30 mL比例、内毒素血症比例、NIHSS评分及颅内压均显著高于非CMODS组,GCS显著低于非CMODS组,住院时间显著更长(均P<0.05)。
- 列线图模型C指数0.804(95%CI 0.768-0.841),校准曲线接近理想曲线、拟合良好(P>0.05)。
- ROC曲线AUC 0.855(95%CI 0.776-0.935),预测CMODS敏感度76.92%、特异度86.57%。
适用人群与场景
- 2019年1月至2023年6月皖南康复医院收治的急性脑出血患者93例,按住院期间是否发生CMODS分为CMODS组26例与非CMODS组67例。
诊断标准与关键阈值
- 以住院期间是否发生脑源性多器官功能障碍综合征(CMODS)分组。
- CMODS组APACHEⅡ≥35分比例显著高于非CMODS组(P<0.05)。
- CMODS组脑出血量≥30 mL比例显著高于非CMODS组(P<0.05)。
研究设计与关键结果
- 研究类型:回顾性分析;样本量93例;人群:急性脑出血患者;CMODS组26例,非CMODS组67例。
- CMODS组较非CMODS组:糖尿病比例、APACHEⅡ≥35分比例、脑出血量≥30 mL比例、内毒素血症比例、NIHSS评分及颅内压显著更高,GCS显著更低,住院时间显著更长(均P<0.05)。
- 多因素Logistic回归:糖尿病OR=3.615(95%CI 1.487-8.785,P=0.000)、APACHEⅡ评分OR=4.697(95%CI 1.933-11.416,P=0.000)、内毒素血症OR=4.577(95%CI 1.883-11.123,P=0.000)、脑出血量OR=4.039(95%CI 1.662-9.816,P=0.000)为CMODS危险因素。
- 以上述危险因素为预测变量建立列线图模型,验证C指数0.804(95%CI 0.768-0.841),校准曲线接近理想曲线、拟合良好(P>0.05)。
- ROC曲线:列线图模型预测CMODS的敏感度76.92%、特异度86.57%,AUC 0.855(95%CI 0.776-0.935)。
证据类型、等级与局限
- 证据类型:回顾性分析。
- 局限:原文未报告局限性。
对床旁实践的改变
- 作者认为,基于糖尿病、APACHEⅡ评分、内毒素血症和脑出血量建立的风险预测模型对评估急性脑出血患者CMODS风险有效。
提炼说明
- 本条依据论文摘要整理;摘要未涉及「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开
OBJECTIVE: To construct and validate a predictive model for the risk of cerebrogenic multiple organ dysfunction syndrome (CMODS) in patients with acute cerebral hemorrhage.
METHODS: Clinical data of 93 patients with acute cerebral hemorrhage admitted to Wannan Rehabilitation Hospital from January 2019 to June 2023 were retrospectively analyzed. Data included baseline information, disease severity score, laboratory examination indicators, cerebral hemorrhage status, treatment status, etc. Patients were divided into CMODS group and non-CMODS group according to whether CMODS occurred during hospitalization. The clinical data of the two groups were compared. Multivariate Logistic regression was used to analyze the risk factors of CMODS in patients with acute cerebral hemorrhage. A nomogram model was constructed to predict the risk of CMODS in patients with acute cerebral hemorrhage, and the model was validated. Receiver operator characteristic curve (ROC curve) was used to evaluate the predictive efficiency of nomogram model for CMODS in patients with acute cerebral hemorrhage.
RESULTS: A total of 93 patients with acute cerebral hemorrhage were enrolled, including 26 patients in CMODS group and 67 patients in non-CMODS group. Compared with the non-CMODS group, the ratio of diabetes, acute physiological and chronic health evaluation II (APACHE II)≥35 score, cerebral hemorrhage volume ≥30 mL, endotoxemia, and national institutes of health stroke scale (NIHSS) and intracranial pressure of patients in the CMODS group were significantly higher, while the Glasgow coma score (GCS) was significantly lower and the length of hospital stay was significantly longer, with statistically significant differences (all P < 0.05). Multivariate Logistic regression analysis showed that diabetes mellitus [odds ratio (OR) = 3.615, 95% confidence interval (95%CI) was 1.487-8.785, P = 0.000], APACHE II score (OR = 4.697, 95%CI was 1.933-11.416, P = 0.000), endotoxemia (OR = 4.577, 95%CI was 1.883-11.123, P = 0.000), and cerebral hemorrhage volume (OR = 4.039, 95%CI was 1.662-9.816, P = 0.000) were the risk factors for CMODS in patients with acute cerebral hemorrhage. Taking the above risk factors as predictive variables, a nomogram prediction model was established. The verification results of the nomogram model showed that the C index was 0.804 (95%CI was 0.768-0.841), and the calibration curve was close to the ideal curve with good fit (P > 0.05). ROC curve results showed that the sensitivity and specificity of the nomogram model in predicting CMODS in patients with acute cerebral hemorrhage were 76.92%, 86.57%, respectively, and the area under the ROC curve (AUC) was 0.855 (95%CI was 0.776-0.935).
CONCLUSIONS: Diabetes mellitus, APACHE II score, endotoxemia and intracerebral hemorrhage are risk factors for CMODS in patients with acute cerebral hemorrhage. The risk prediction model based on these risk factors is effective in evaluating the risk of CMODS in patients with acute cerebral hemorrhage.
参考来源
- Wu H, Zhang H, Guo C, et al. [Establishment of a risk prediction model for cerebrogenic multiple organ dysfunction syndrome in patients with acute cerebral hemorrhage]. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2024. PMID: 39380519. doi:10.3760/cma.j.cn121430-20240202-00109
供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。